(50-5) 09 * << * >> * Russian * English * Content * All Issues

Algorithm for segmentation and analysis of pigment epithelial detachment
R.T. Samigullin 1, N.Yu. Ilyasova 1, 2, D.V. Kirsh 1, 2, Е.Yu. Zubkova 3, E.A. Zamytskiy 4, N.S. Demin 1, 2

1 Samara National Research University, Moskovskoye Shosse 34, Samara, 443086, Russia;
2 Image Processing Systems Institute, NRC "Kurchatov Institute",
Molodogvardeyskaya St. 151, Samara, 443001, Russia;
3 Samara Regional Clinical Ophthalmological Hospital named after T.I. Eroshevky, Samara,
443068, Russia, Samara, st. Novo-Sadovaya, 158;
4 Samara State Medical University, Chapaevskaya St. 89, Samara, 443099, Russia

  Full text (PDF)

DOI: 10.18287/COJ2000

Article ID: 2000

Language: Russian

Abstract:
The paper presents an algorithm for automatic segmentation of pigment epithelial detachment (PED) in optical coherence tomography (OCT) images, followed by a quantitative assessment of its morphometric characteristics. A convolutional neural network with a U-Net architecture, trained on expert-annotated OCT-B scans, was used to highlight the PED area. Segmentation accuracy was assessed using the Dice Score and Intersection over Union (IoU) metrics and showed high correspondence with the reference annotation (Dice = 0.8614; IoU = 0.7590). Based on the segmented data, a three-dimensional reconstruction of the PED was performed and the linear dimensions, area, and volume were calculated, taking into account the metrological parameters of the tomograph. The error assessment showed that the total relative uncertainty of the volume measurements did not exceed 2–3%. The proposed approach provides an objective and reproducible quantitative assessment of PED and can be used in clinical and research tasks.

Keywords:
pigment epithelial detachment (PED), optical coherence tomography (OCT), image segmentation, deep learning, quantitative assessment, retinal biomarkers, morphology analysis.

Acknowledgements:
The work was carried out within the state assignment of NRC “Kurchatov Institute” (theoretical part) and within the state assignment of Samara University, theme FSSS-2024-0014 (experiments and software implementation).

Citation:
Samigullin RT, Ilyasova NYu, Kirsh DV, Zubkova EYu, Zamytskiy EA, Demin NS. Algorithm for segmentation and analysis of pigment epithelial detachment. Computer Optics 2026; 50(5): 2000. doi: 10.18287/COJ2000.

References:

  1. Khanani AM, Guymer RH, Kaiser PK, Nishijo K, Wolf A, Gekkieva M, Heier JS. TENAYA and LUCERNE: Two-year results from the phase 3 neovascular age-related macular degeneration trials of faricimab with treat-and-extend dosing in year 2. Ophthalmology 2024; 131(8): 914-926. DOI: 10.1016/j.ophtha.2024.02.014.
  2. Leitritz MA, Gelisken F, Inhoffen W, Voelker M, Helb HM, Pauleikhoff D, Bartz-Schmidt KU, Schmitz-Valckenberg S. Can the risk of retinal pigment epithelium tears after bevacizumab treatment be predicted? An optical coherence tomography study. Eye 2008; 22: 1504-1507. DOI: 10.1038/eye.2008.145.
  3. Nawash B, Lakshmi K, Al-Sweedan SA, Alqudah A. Prognostic optical coherence tomography biomarkers in neovascular age-related macular degeneration. Journal of Clinical Medicine 2023; 12(9): 3049. DOI: 10.3390/jcm12093049.
  4. Rispoli M, Romano MR, Sacconi R, Marchese A, Parrulli F, Bandello F. Practical guidance for imaging biomarkers in exudative age-related macular degeneration. Survey of Ophthalmology 2023; 68(4): 615-627. DOI: 10.1016/j.survophthal.2023.02.004.
  5. Ul Kadir SM, Shamsuddin S, Azlan CAN, Yap CY, Tajul Arifin K. Artificial intelligence in ophthalmology: An overview. Ophthalmology Research 2022; 5(1). DOI: 10.33425/2639-9482.1024.
  6. Sivaguru R, Venkatesh D, Kumar A. Applications of artificial intelligence in the medical field: A survey. International Research Journal of Advanced Engineering Hub 2024; 2(3): 407-415. DOI: 10.47392/IRJAEH.2024.0060.
  7. Pedanova EK, Doga AV. Treatment of polypoid choroidal vasculopathy: Photodynamic therapy, anti-VEGF therapy, or their combination? A review [In Russian]. Ophthalmology in Russia 2019; 16(2): 151-158. DOI: 10.18008/1816-5095-2019-2-151-158.
  8. Saeidian J, Shafabakhsh R, Fathi A, Moeini A, Javanmard SH. Segmentation of choroidal area in optical coherence tomography images using a transfer learning-based conventional neural network: A focus on diabetic retinopathy and a literature review. BMC Medical Imaging 2024; 24(1): 281. DOI: 10.1186/s12880-024-01459-2.
  9. Marciniak T, Stankiewicz A, Zaradzki P. Neural networks application for accurate retina vessel segmentation from OCT fundus reconstruction. Sensors 2023; 23(4): 1870. DOI: 10.3390/s23041870.
  10. Leandro J, Nunes S, Alves D, Santos T, Leite J, Marques IP, Figueira J, Pires I. OCT-based deep-learning models for the identification of retinal key signs. Scientific Reports 2023; 13(1): 14628. DOI: 10.1038/s41598-023-41362-4.
  11. Zhang H, Jia Y, Liu Y, Chen H. Retinal OCT image segmentation with deep learning: A review of advances, datasets, and evaluation metrics. Computerized Medical Imaging and Graphics 2025; 123: 102539. DOI: 10.1016/j.compmedimag.2025.102539.
  12. Eckardt F, Menghini M, Birtel J, Weber RG, Herrmann P, Pfau M, Holz FG, den Hollander AI, Charbel Issa P. Deep learning-based retinal layer segmentation in optical coherence tomography scans of patients with inherited retinal diseases. Klinische Monatsblätter für Augenheilkunde 2025; 242(9): 903-911. DOI: 10.1055/a-2227-3742.
  13. Lin M, Chen Q, Liu J, Ma Y, Yu W. Recent advanced deep learning architectures for retinal fluid segmentation on optical coherence tomography images. Sensors 2022; 22(8): 3055. DOI: 10.3390/s22083055.
  14. Pattathil N, Zhao JZL, Sam-Oyerinde O, Felfeli T. Adherence of randomised controlled trials using artificial intelligence in ophthalmology to CONSORT-AI guidelines: A systematic review and critical appraisal. BMJ Health & Care Informatics 2023; 30(1): e100757. DOI: 10.1136/bmjhci-2023-100757.
  15. Balas M, Micieli JA, Wong JCY. Integrating AI with tele-ophthalmology in Canada: A review. Canadian Journal of Ophthalmology 2025; 60(3): e337-e343. DOI: 10.1016/j.jcjo.2024.08.013.
  16. Kugelman J, Zhang J, Abou Shousha M, Feuer W, Kishor K, Baniasadi N, Wang J, Lee CS, Lee AY. A comparison of deep learning U-Net architectures for posterior segment OCT retinal layer segmentation. Scientific Reports 2022; 12(1): 14888. DOI: 10.1038/s41598-022-18646-2.
  17. Wani MY, Bhat JI, Mir AH. Automated retinal blood vessel segmentation via U-Net and VGG-based models. Spectrum of Engineering Sciences 2025; 3(4): 490-498.
  18. Karn PK, Abdulla WH. Advancing ocular imaging: A hybrid attention mechanism-based U-Net model for precise segmentation of sub-retinal layers in OCT images. Bioengineering 2024; 11(3): 240. DOI: 10.3390/bioengineering11030240.
  19. Shinde R. Glaucoma detection in retinal fundus images using U-Net and supervised machine learning algorithms. Intelligent-Based Medicine 2021; 5: 100038. DOI: 10.1016/j.ibmed.2021.100038.
  20. Pavani PG, Biswal B, Gandhi TK, Kota AR. Robust semantic segmentation of retinal fluids from SD-OCT images using FAM-U-Net. Biomedical Signal Processing and Control 2024; 87: 105481. DOI: 10.1016/j.bspc.2023.105481.
  21. Michl M, Schlegl T, Waldstein SM, Bogunovic H, Podkowinski D, Gerendas BS, Sadeghipour A, Schmidt-Erfurth U. Automated quantification of macular fluid in retinal diseases and their response to anti-VEGF therapy. British Journal of Ophthalmology 2022; 106(1): 113-120. DOI: 10.1136/bjophthalmol-2020-317764.
  22. Aksenova LE, Fedorenko SV, Kurenkov AS, Zaytsev AV, Zheltov SYu, Khoroshilov AV. Development and testing of an automated system for retinal OCT image analysis [In Russian]. Doklady of the Russian Academy of Sciences. Mathematics, Informatics, Control Processes 2023; 514(2): 169-176.
  23. Ibragimova RR, Shchuko AG, Zhidkova SA, Pashkina AV, Khoroshilov AV. Algorithm for segmentation of macular OCT images for analysis of patients with age-related macular degeneration [In Russian]. Bulletin of Russian State Medical University 2022; (6): 89-96.

151, Molodogvardeiskaya str., Samara, 443001, Russia; E-mail: journal@computeroptics.ru; Tel: +7 (846) 242-41-24 (Executive secretary), +7 (846) 332-56-22 (Issuing editor), Fax: +7 (846) 332-56-20